You Paid for the Whole Board and Costed Only the Part You Used
· 7 min read · Written by Faceela Research & Editorial Team
Reviewed by Ahmed Hassan Algammal — Founder and Enterprise Systems Consultant
A board arrives on a pallet and you pay for all of it. The saw takes the parts the drawing asked for. What is left leans against a wall until somebody needs the space.
The job is costed on the parts. The board was bought whole. The difference between those two sentences is the entire subject of this article, and in a trade where materials are 55–70% of the cost of a job, it is not a housekeeping matter. It is where the margin is decided.
Most workshops know this and treat it as unmeasurable. It is not unmeasurable. It is unmeasured, which is a different word with a different remedy.
The cutting list is a commercial document that everyone treats as a workshop one
Follow what the cutting list actually is. The drawing says what the client gets. The cutting list says what has to be cut to produce it. From that list comes the quantity of board to buy, the hardware to order, the hours to plan and the price the job was quoted at.
Which makes it the bridge between the drawing and the purchase — the single document where a design becomes money.
And in most workshops it is a spreadsheet on one machine, maintained by one person, emailed as an attachment, and revised by overwriting. When the drawing changes, there is a version on the saw and a version in the office and no rule about which one the purchase order was raised from.
The shape of it
One board you paid for in full, four places it ends up
One sheet, bought whole
The parts on the drawing
The offcut worth keeping
The offcut nobody will use
Kerf, trim and dust
1. The parts on the drawing
The pieces the cutting list asked for, nested onto the board and cut.
What it costs: Nothing extra. This is the only part of the board the job was ever costed for.
Who records it today: The cutting list, and through it the job. This is the part every workshop already has.
2. The offcut worth keeping
A piece long enough and clean enough to be the shelf on somebody else's job next month.
What it costs: Nothing, if it is found again. Full price, if it is not — because the same board gets bought.
Who records it today: Nobody. It is stood against a wall. It has no dimension, no material code and no location.
3. The offcut nobody will use
Too short, wrong grain, damaged edge, or simply the third one this week of a size nothing needs.
What it costs: The board it was cut from, minus the parts. Real money, spent, with no line anywhere.
Who records it today: Nobody. It goes to the rack, then to the skip, and the decision between the two is made by whoever is tidying up.
4. Kerf, trim and dust
Saw kerf, the trim off two edges, the sanding, the strip lost to the clamp.
What it costs: Small per board and structural per year. It is the one everybody knows about.
Who records it today: A flat wastage percentage in the estimate, applied identically to every material and every sheet size.
Why this is a margin question and not a housekeeping one
Materials are 55–70% of the cost of a job in this trade. That makes the split above the largest single thing a fabrication workshop can influence — and three of its four destinations currently reach no job, no report and no quotation.
Theoretical yield and the yield the operator got
Nesting software will tell you what a board can produce. That figure is theoretical yield — the best arrangement of the required parts on the sheet, computed by an optimiser that assumes the parts are interchangeable rectangles and the operator follows the plan.
What the workshop actually got is achieved yield, and it differs for reasons that are all legitimate. A sheet had a defect and the layout was shifted. A part was cut wrong and recut from the same board. The operator worked from a printed layout that was two revisions old. Two jobs were on the saw at once and the cuts were combined by eye, which is often better than the software and is never recorded as anything.
The gap between the two is the only number in this article that tells you something you can act on, because it is the only one that separates the plan was optimistic from the plan was not followed. Almost nobody in the trade holds both numbers, so almost nobody can tell those two apart — and they have completely different fixes.
The offcut is a stock item that nobody will call a stock item
An offcut is either worth keeping or it is not, and the workshop makes that decision every day without ever writing it down.
Consider what "keeping" requires. The piece needs a dimension, a material, a location and a way of being found by somebody who is not the person who cut it. It has none of those. It has a wall.
So one of two things happens. Either it is thrown away, and the board it came from has been bought for parts worth less than the board. Or it is kept, which feels thrifty, and then the same board is bought again next month because nobody searching for a piece of 18mm oak-veneered MDF, 900 by 400, has any way of discovering that one is standing behind the edge bander.
The second outcome is the expensive one, because it costs the price of the board and the floor space and the eventual skip. It is also the one that looks like good practice while it is happening.
There is a size below which none of this is worth doing, and every workshop's number is different. The point is not that offcuts must all be tracked. It is that the threshold should be a decision the business made once, rather than a decision made by whoever is tidying up on a Thursday.
The optimiser that produces an excellent unusable answer
Now the part that makes joinery different from sheet metal.
A nesting optimiser that treats a veneered board as a rectangle will return a mathematically excellent layout in which the grain runs across three doors and along the fourth. It is the best answer to the question it was asked. It is also scrap.
| What the optimiser has to be told | What it produces when nobody tells it |
|---|---|
| Grain direction per part | Doors and panels with the grain running the wrong way on some of them |
| Which parts form a matched set | A book-matched run assembled from four different areas of the log |
| Face side and defect map of the sheet | The knot placed in the middle of the customer-facing door |
| Saw kerf and edge-banding allowance | Parts consistently undersize by the thickness of the blade and the tape |
| Trim allowance on the factory edge | A layout that uses a mill edge nobody would put on a finished piece |
| The minimum offcut size worth keeping | Remnants shaped so that every leftover falls just below usable |
Every row on that table is a piece of knowledge that lives in the head of a machinist, and every row is a cost that lands on the job with no name on it. Machine optimisation without those constraints is not neutral — it produces confident, well-nested, unusable answers faster than a human could produce careful ones.
The wastage factor that is the same number for every material
Ask an estimator what wastage he adds and you will get a percentage. Ask him whether it is the same percentage for a 2440 by 1220 board and for a 3050 by 1300 sheet, for a run of identical wardrobes and for a one-off reception desk, and the answer is almost always yes.
It cannot be right, and everybody knows it cannot be right. A large repetitive job nests beautifully and a small varied one does not. A big sheet gives better yield on big parts and worse yield on small ones. A dark high-gloss laminate has a rejection rate that a melamine white does not.
The flat factor survives because the alternative — a real figure, per material, from history — requires two things a workshop rarely has: the cutting list and the purchase in the same system, and the achieved yield recorded against the job rather than inferred from the stock count at year end.
That is the whole prerequisite. It is not a modelling problem. It is a plumbing problem, and it is the same plumbing that answers which job types consistently lose money.
What to make a vendor show you
On a live system, with a real drawing and a real board, not on slides.
- Take a cutting list from a job and show the material requirement it produces, netted against what is in the store.
- Change the drawing, re-issue the list, and show what happens to a purchase order already raised against the old one.
- Record what was actually cut against what the layout said, on one job, and show the difference.
- Book an offcut into stock with a dimension, a material and a location, and then find it from a later job's requirement.
- Charge the waste on a board to the job that caused it, and show it on that job's cost sheet as a separate line from the material consumed.
- Show the wastage percentage the system derived from history, per material, against the flat number the estimate used.
- Show grain direction and matched-set constraints travelling from the drawing to the cutting list without anyone retyping them.
Item 3 is the one that decides it. Everything else on the list is a feature; that one is a comparison, and a comparison is the only thing that can tell you whether a plan was wrong or merely ignored. A vendor who can show planned against achieved on one real job has a system with a memory. A vendor who cannot has a system that will produce beautiful layouts and learn nothing from any of them.
Item 4 is the quiet one. Most systems can hold an offcut as a stock line. Very few can find it from a later requirement, which is the only reason to hold it at all.
The short version
Yield is not a workshop statistic in this trade. With materials at 55–70% of job cost, it is the largest single lever a fabrication business has, and it is the one most consistently left to habit.
The board divides into four destinations and only one of them reaches a job. The offcut has no name, no size and no location, so it is bought twice or thrown away once. The optimiser gives a perfect answer to a question that left out grain. And the wastage factor is a single number standing in for every material in the building.
None of that needs a better saw. It needs the cutting list, the purchase and the actual cut to exist in the same place, so that the difference between what was planned and what was achieved is a number somebody can look at on a Monday.
The estimate that this all feeds is estimating joinery jobs, the cost sheet it should land on is three costs that never reach the job, and the reason the store cannot see any of it is why your stock figure is wrong. What it looks like built as one system is on the joinery, steel and fabrication page, and the system is Flow.